Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

semafor.com

Post reply on HN
Print view View original

691–700 of 873 posts

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#691

Post by myko »

polski-g wrote: post #445
sethammons wrote: post #442

Earlier quoted context omitted.

> As described in court filings, from at least December 2022 through December 2024, Mazulina conspired with Russian freight forwarding companies and others to unlawfully ship controlled items, including industrial oil and gas equipment, from the United States to Russia, through intermediary countries. At one point, in June 2023, Mazulina told colleagues that her clients were paying through bank accounts in third part…

Anthropic doesn't have the right to release weapons technologies on the open Internet, that's the point. Lockheed doesn't whine that they aren't allowed to sell their products for a $20/mo subscription to everyone with a pulse.

> doesn't have the right to release weapons technologies

true but that isn't what we're talking about here

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#692

Post by djeastm »

daseiner1 wrote: post #556
amarant wrote: post #554

Dear anthropic, please consider moving your entire business to Canada or Europe where you will be allowed to conduct your business without vindictive and Kafkaesque government interference. I would like to have access to your models too!

famously friendly business environments europe & canada lol

I think Anthropic could manage

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#693

Post by 4lx87 »

andsoitis wrote: post #131
4lx87 wrote: post #102

Earlier quoted context omitted.

We don’t have to guess. Democrats had a pragmatic policy under Biden — which was rescinded by Trump when he took office. https://bidenwhitehouse.archives.gov/briefing-room/statement...

> Democrats had a pragmatic policy under Biden How well does it stand up to Mythos?

Hard to imagine it would be worse than Republicans, who don’t have any policy except for the whims of a man child.

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#694

Post by K0balt »

The real reason, afaik, that the US is trying to restrict access to SOTA models is that a very large component of USA tailored access and surveillance relies on exploits and weaknesses that these models will easily detect.

Thus, it really is an export control issue, but it has nothing to do with offensive capabilities. Offensive capabilities always exist, but pervasive defensibility would upset the asymmetric advantage that attackers, especially the USA, currently have.

There are now Asian models coming , optimized focused on cybersecurity defense at a high level, so I suspect this will be a relatively moot point soon.

LLMs are not great at creating exploits, but they are really good at detecting them. That asymmetry alone is enough to destroy the “offensive capabilities” narrative.

Yes, mythos can find exploitable bugs, even write bench exploits. But real exploits require a good dose of human psychology, and most of the tools needed are off the shelf available anyway. You still need a real cybersecurity expert to effectively weaponize a zero day into a deployable exploit.

But an LLM can inspect payloads, packages, and blobs en masse and find those exploits in a way that was wholly impractical before, so the asymmetric attack advantage is dissolved by strong LLMs.

The USA is trying to protect its cyberwarfare advantage, not protect against attackers. The exact opposite, actually. Porous security is a huge advantage to technologically advanced state actors.

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#695

Post by the_gastropod »

Alupis wrote: post #220
SpicyLemonZest wrote: post #214

Earlier quoted context omitted.

Again, you're seeing this as some kind of clever invariant of politics, but it's not. People are blind to politicians on their side because Donald Trump has spent the past 10 years working hard to make sure that everyone in politics is constantly enraged at the other side. You're being "punished to hell", in your words, because you're blaming Trump's passionate embrace of hatred-based politics on his opponents.

> you're seeing this as some kind of clever invariant of politics, but it's not. People are blind to politicians on their side because Donald Trump Yes, sure, it is I who is blind...

lol. Yes. You couldn’t possibly be blind! You’re a both-sides guy. And, as everyone knows, both-sides guys are rational and grounded in reality.

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#696

Post by lantry »

t0mas88 wrote: post #438
willsmith72 wrote: post #276

Earlier quoted context omitted.

Oh you thought this was a competitive free market?

I thought "Free market!!1!" was basic US policy. But apparently you can't do any consumer protection laws, but you can regulate AI access to only your wealthy friends / campaign sponsors.

"there are in-groups that the law protects but does not bind, and out-groups that the law binds but does not protect"

i.e. freedom for me but not for thee

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#697

Post by igregoryca »

pdimitar wrote: post #475
didibus wrote: post #320

Earlier quoted context omitted.

> I don't have too much unknowns in my work, I simply have way too much to do. Interesting, I'm curious what work you do? My software engineering career has never been in that situation, it's always so much ambiguity and unknown that trumps everything.

Fair question, and I was vague just so as not to balloon the comment. I work in a financial startup. The codebase is a mess and very much spaghettified. One rework that forced us to migrate our data model from 1:1 users loans to M:N (many-to-many) took two months and touched ~40% of the codebase... multiple times. Huge churn. And it just crossed two months of work, even though it's now in its very final phases. I kno…

I know this is a tangent, but

> Introduce and enforce structs for passing context and input shapes around. So as to stop fighting with NULLs, lack of keys in maps and other maddening cases that inflate your coding lines for no other reason than programming languages not having higher-order constructs on well-researched and mostly resolved computer science problems

Amen to that.

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#698

Post by polski-g »

myko wrote: post #691
polski-g wrote: post #445

Earlier quoted context omitted.

Anthropic doesn't have the right to release weapons technologies on the open Internet, that's the point. Lockheed doesn't whine that they aren't allowed to sell their products for a $20/mo subscription to everyone with a pulse.

> doesn't have the right to release weapons technologies true but that isn't what we're talking about here

Then why did Dario say that their product was like nuclear material? https://coincentral.com/anthropic-ceo-says-ai-could-be-as-da...

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#699

Post by _3u10 »

someguyornotidk wrote: post #558

If this becomes the norm, what incentive does the rest of the world have to keep their markets open to the US? If US companies have a large unfair advantage such that domestic competitors are no longer able to compete, then wouldn't it make sense for governments around the world to ban or tariff US products and services? If I was responsible for national economic policy, I would place this at the top of my non-emerge…

They don’t, they wouldn’t need tariffs if they did.

America mostly produces cheaper ag commodities than the EU but more expensive than South America. Deepseek is already a better search engine than Google. (Not sure if it does Google searches)

Travel the world it’s not American companies gaining market share. I would especially recommend trying Chinese AI or riding in a BYD car and judging for yourself.

Re: U.S. allows Anthropic to release Mythos AI to ‘trusted’ US organizations

#700

Post by brokencode »

drdaeman wrote: post #282
brokencode wrote: post #98

Earlier quoted context omitted.

I think they’d try to get something through Congress to regulate the industry in a rules-based way. The current admin flies by the seat of their pants and at least creates the perception of political decision making.

What rules, though? Bernie and AOC (which aren't DNC mainstream, but prominent) had just pushed for a moratorium on "AI data centers" with a definition that includes "that are used for the development or operation of AI models at scale" (trivially sidesteppable by "we build this GPU farm to sell to whoever bids for compute" - which is actually true), plus a bunch of fancy extras bundled in like "The government must r…

The whole point of Congress is to get a bunch of people with different ideas and hash them out. Ideas like this are an input, not an output, of Congress.
Post reply on HN